Uncertainties, short deadlines, missed alerts… predicting ocean behaviour remains a major scientific challenge, while some regions of the globe offer no respite. In the tropical Atlantic in particular, no warning was given to the population about the El Niño events of 2021. In response to this situation, a group of scientists from the University of Bergen, the CMCC, the LEGOS and the University of Cape Town have decided to use artificial intelligence to test a new method for predicting extreme events in this key area of the tropical Atlantic Ocean.
By Laurie Henry
Extreme weather events in the tropical Atlantic, such as El Niño or La Niña in the Atlantic or Benguela, cause heavy rainfall in West Africa, disrupt marine ecosystems and influence global climate change. However, predicting these events on a seasonal scale remains difficult.
Unlike the Pacific, these regions exhibit complex variability that is difficult to model using traditional approaches. New research published in April 2025 in Science Advances uses artificial intelligence to test a new method for forecasting extreme events in this key area of the ocean.
